Table Of Content

EARLY SILURIAN SINACANTHS (CHONDRICHTHYES) FROM CHINA by ZHU MIN Abstract.HistologicalstudyofnewspecimensofsinacanthfinspinesfromtheLowerSilurianofthenorth- westernmarginoftheTarimBasin(Xinjiang,China)showsthattheyhavethesamehistologyasthefinspines ofchondrichthyans.Onthisbasisitisarguedthatsinacanthsareoneoftheoldestknownchondrichthyans, ratherthanacanthodians,andtheirspinesaretheoldestknownsharkfinspines.Previousstudiesonsinacanths are critically reviewed. The family Sinacanthidae is erected to include Sinacanthus and its relatives with more than 15 fin spine ridges per side. It is suggested that Sinacanthusfancunensis is synonymous with S. wuchangensis.Anewsinacanthgenusandspecies, Tarimacanthusbachuensis,fromtheLowerSilurianof TarimandSouthChina,iserected. ThesinacanthsareamiddlePalaeozoicfishgroup,exemplifiedbySinacanthuswuchangensiswhich waserectedbyP’an(1959, 1964)forisolatedfinspinesfromtheGuodingshanFormation(Silurian: Wenlock) ofWuhan, China (Text-fig. 1), and referred originally to Acanthodidae. Liu (1973) reportedSinacanthusinNingguo,Anhui,China(Text-fig.1).Sincethen,largenumbersofsinacanth specimens,allofSilurianage,havebeenfoundineightprovincesofSouthChina(P’anetal. 1975; Li 1980;Pan1986a, 19866;Zeng1988;Text-fig. 1).TheyareakeyelementoftheendemicSilurian vertebrate fauna of South China, and have also been significant in regional stratigraphical correlation(Pan 19866). Turner (1986) assigned to ‘cf. Sinacanthus' some fin spines from the Lower Devonian of Australia, previously reported by Chapman (1917) and Talent and Spencer-Jones (1963). She indicatedsomedoubtastotheiracanthodianaffinity,andsuggestedthattheymayhavebelonged to a shark akin to Antarctilamna Young, 1982. At the same time, fin spines from the Upper Silurian-MiddleDevonianofBoliviawereconsideredtoresembleSinacanthus(JanvierandSuarez- Riglos1986;Gagnieretal. 1988),andGagnieretal.(1988)erectedSinacanthusboliviensis.Gagnier etal.(1988)placedSinacanthusinAcanthodii(orderandfamilyundetermined),butsuggestedthat Sinacanthusmightbeachondrichthyanratherthananacanthodian. Becauseallofthefinspines were disarticulated and preserved mainly as external or internal moulds, there was no sound evidence to distinguish them from either chondrichthyans or acanthodians, and the systematic positionofSinacanthusandrelatedformsremainedunclear.Inarecentstudy,Liu(1995,footnote on p.94) doubted the presence of Sinacanthus in Australia and Bolivia, and returned to the traditionalclassificationofsinacanthsasacanthodians.Ihavehadtheopportunitytoexaminethe Bolivianfinspines,currentlyhousedinParis(MuseumNationald’HistoireNaturelle),andamable to confirm that both sinacanths and acanthodians are present in Bolivia. As to sinacanths in Australia, I follow Turner’s (1986) proposal, based on the material illustrated by Talent and Spencer-Jones(1963)whichshowssimilaritiestotheChinesesinacanths. ThesinacanthmaterialreportedinthisstudywascollectedfromtheSilurianofthenorth-western marginoftheTarimBasinin1992bytheauthorandhiscolleagues(WangJunqingandLiuShifan; Text-fig. 1, localities 11 and 12). Abundant fin spines, some galeaspids and large numbers of vertebratemicroremainsincludingdermalscale-unitsofgaleaspidsandscalesofchondrichthyans indicateadiversevertebratefaunasimilartotheLlandovery-WenlockvertebratefaunaofSouth China.ThisfaunawasfirstreportedfromTarimbyWangetal.(1988),andfurtherinvestigatedby [Palaeontology,Vol.41,Part1,1998,pp.157-171,1pi.) ©ThePalaeontologicalAssociation 158 PALAEONTOLOGY,VOLUME41 text-fig. 1. SincanthlocalitiesinChina.Dottedlinesrepresentprovinceboundaries. 1,NanjingandWuxi, Jiangsu;2,Ningguo,Anhui,andChangxing,Zhejiang;3,Chaoxian,Anhui;4,Jingshan,Hubei;5,Wuhan, Hubei;6,Ruichang,Jiangxi;7,Lixian,Hunan;8,Dayong,Hunan;9,Xiushan,Sichuan;10,Kaili,Guizhou; 11,Kalpin,Xinjiang; 12,Bachu,Xinjiang. Liu(1995)andWangetal. (1996). ThediscoveryofsinacanthsinTarimsupportsotherevidence forbiogeographicalaffinitybetweentheSouthChinaandTarimblocksintheSilurian(Liu 1993, 1995;Wangetal. 1996).Thenewsinacanthspinescanbeusedtodeterminethesystematicposition ofthisgroupbecause,althoughdisarticulated,likethosefromSouthChina,theyarecomposedof wellpreservedhardtissuesandaresuitableforhistologicalresearch. Resolutionofthedebatablesystematicpositionofsinacanths,usinghistologicaldetails,willalso helptoclassifythevariousfinspinespecies.InSouthChina,fivespecies(P’an1959,1964;Liu1973; P’anetal. 1975),aswellasseveralunnamedforms(Zeng 1988),havebeenreferredtosinacanths. However,differenttypesoffinspinefoundinthesamebed,suchasSinacanthuswuchangensisand S.triangulatus,mightrepresentdifferenttaxaiftheyarechondrichthyans,ormaybebetterregarded as from the same fish ifthey are acanthodians. For this reason, the systematic revision ofthe sinacanthsgivenbelowfollowsthediscussionofthesystematicaffinityofsinacanths. Institutional abbreviations. GM.V (Geological Museum of China, Beijing); HV (Regional Geological SurveyingTeam,BureauofGeologyandMineralResourcesofHunanProvince,Xiangtan);IVPP.V(Institute ofVertebratePaleontologyandPaleoanthropology,Beijing);VF(InstituteofGeology,ChineseAcademyof GeologicalSciences,Beijing). ) ZHU: EARLYSILURIANSINACANTHS 159 HISTORICAL REVIEW OF STUDIES OF SINACANTH SinacanthuswaserectedbyP’an(1959),withonespecies(<S.wuchangensis)describedbrieflyonthe basisoftwoincompletefinspines(P’an1959,pi.4,figs3-4).However,noholotypewasdesignated, sothetwospecimensmustbeconsideredassyntypes.Togetherwiththesetwofinspines,afragment ofthe headshield ofHanyangaspis (P’an et al. 1975) was found. This was first referred to the antiarchs by P’an (1959), but is in fact the first galeaspid to be figured from China (P’an 1959, p1i9.559,aficgc.o7r;dcifn.gLitout1h9e65I)n.teIrnnaadtdiiotniaoln,CotdheecoiftaZtoioolnodgaitcealofNSoimneanccalnatthuusrean(dRiSd.eweutcahla.n1g9e8n5s,isArsthioclueld22b)e, ratherthan 1957(P’an 1959, 1964; P’anetal. 1975;Turner 1986; Gagnieretal. 1988; Liu 1993, 1995). P’an(1964)publishedamoreformal study ofSinacanthusand S. wuchangensis, based onnew materialfromthetypelocality(Wuhan, China). Healsodesignatedaholotype(P’an 1964,pi. 1, fig. la-b; see also P’an 1959, pi. 4, fig. 3) and a paratype (P’an 1964, pi. 1, fig. 2a-b) of S. wuchangensis. However, since both holotype and paratype should be fixed in the original publication (Rideetal. 1985, Article 73), the holotype ofP’an (1964) should be regarded as the lectotypeofS. wuchangensis, andhisparatypeastheparalectotype. A second species ofSinacanthus (S.fancunensis) was described from Ningguo, Anhui, China (Text-fig. 1) by Liu (1973), but it is argued below that this species is synonymous with S. wuchangensis. P’an et al. (1975) described a third species ofSinacanthus (S. triangulatus) and a second sinacanth genus (Neosinacanthus from the type locality ofS. wuchangensis. Since then, sinacanthshavebeenfoundtobewidelydistributedinSouthChina(Lietal. 1978;Xia 1978;Pan 1984, 1986a, 19866;PanandDineley 1988; Zeng 1988; Text-fig. 1). RegardingtheageofsinacanthsfromSouthChina,P’an(1959)firstproposedamidDevonian age for S. wuchangensis, but later (1964) modified this opinion to Devonian or older, while Liu (1973)regardedtheageofS.fancunensisasEarlyDevonian.P’anetal.(1975)alsoadoptedanEarly Devonian age for S. wuchangensis, S. triangulatus and Neosinacanthusplanispinatus. The Early DevonianageofSinacanthusandNeosinacanthushasbeengenerallyaccepted(Hou 1978;P’anet al. 1978;Xia1978),althoughsomehavearguedforanolder,midSilurianage(Lietal. 1978).More recently,evidencefrominvertebratefossilsandstratigraphicalsequenceshassupportedaSilurian ageforthesinacanths(Li 1980; Wangetal. 1980; Lin 1982; Yangand Rong 1982), aviewnow supported byprevious proponents ofan Early Devonian age (Pan 1986a, 19866; Liu 1995). To summarize, sinacanths are recorded onlyfrom the Llandovery toWenlock ofSouthChina (Pan 1986a, 19866; Pan and Dineley 1988; Zeng 1988; Liu 1995). Reports by Turner (1986), who assumedaSilurian-EarlyDevonianageforSinacanthusinSouthChina,andJanvierandSuarez- Riglos(1986),whostatethatSinacanthusisknownfromtheLowerandMiddleDevonianofChina, areincorrect. SINACANTH HISTOLOGY Itisgenerallydifficulttodistinguishbetweendisarticulatedridgedfinspinesofacanthodiansand chondrichthyans. However, they exhibit obvious differences in histological structure. In acanthodians,thefinspinesusuallyconsistofthreelayers:superficial,middle,andbasal. Insome maturefinspines,acentralcavityosteonordenteonmayfillinthecentralcavity(Denison 1979). Thesuperficiallayer(alsoreferredtoasthesculpturelayer)formstheridgesofthefinspine.This layer,composedoforthodentineormesodentine(0rvig 1967),isdistinguishedhistologicallyfrom theunderlyingmiddlelayer,whichisformedofeithercellularboneortrabeculardentine (0rvig 1967). When present, the basal layer lining the central cavity is composed ofcellular bone or dentine. Enamel or enameloid tissue has not been identified in the fin spines ofacanthodians. Amongst chondrichthyans, only some fossil sharks, such as Ctenacanthus Maisey, 1981, AntarctilamnaYoung, 1982and TristychiusZangerl, 1981,haveridgedfinspines.Thesefinspines consistoftwolayersand,insomespecimens,athinenamellayeronthesurface.Internally,athin basallayerlinesthecentralcavity,asinacanthodians.Thisinnerlayerisformedoflamellardentine 160 PALAEONTOLOGY,VOLUME41 (Maisey1975).Theouterlayer,composedoftrabeculardentine,formstheridgesandtrunkofthe finspine.Thus,thehardtissueformingthefinspineridgesinchondrichthyansistrabeculardentine, whereasinacanthodiansitisorthodentineormesodentine.Therefore,theidentificationofthehard tissueformingthefinspine,especiallythat oftheridges, isessentialforthedeterminationofthe systematicaffinityofdisarticulatedfinspines. Fourthinsectionsshowthehistologicalstructureofsinacanthfinspines.Twoarecrosssections offin spines (Text-figs 3f,4c), which are oval in form and very compressed. IVPP.V11249 was sectionednearthebase(a-a',Text-fig.3d),wheretheposteriorwallisopen,andhasawidth/height indexofnearly600(Text-fig. 3f).IVPP.V11250wassectionedneartheapexofthefinspine(a-a', Text-fig.4b),whereitiscompletelyenclosed,andhasawidth/heightindexofabout500(Text-fig. 4c).Inbothexamples,thespinewallisfairlythin,andthereisarelativelylargecentralcavity(c.cav, Text-figs3f,4c).Athincompactlayer,consistingoflamellardentine(lam,Text-figs3f,4c;PI. 1, fig. 1), lines the central cavity. It resembles the inner dentinous layer offin spines from spinate sharks(Maisey 1975, 1979;Young 1982),orthebasallayerinthefinspinesofsomeacanthodians (Gross 1947;Denison 1979).Coveringtheinnerlamellarlayerisazoneoftrabeculardentine(tra. Text-figs3f,4c;PI. 1,fig. 1),whichformsthemainbodyofthefinspine,andispenetratedbymany vascularcanals.Thisouterlayerextendsintotheridges,whosetissueisthushomogeneouswiththat immediatelybeneaththem.Thetrabeculardentineintheridgesisalsopiercedbyvascularcanals, although they areless numerous thanthosein themainbody ofthefin spine. No enameloid or enamel tissue was found on the surface of the ridges, or in the grooves between the ridges. Enameloidorenameltissueispresentinsome,butnotall,finspinesofspinatesharks(Maisey1981, 1982;Young 1982). SYSTEMATIC POSITION OF SINACANTHS Evidencesupportingthechondrichthyanaffinityofsinacanthsisasfollows. 1. Thetissueinthefinspineridgesistrabeculardentineandisthesameasthetissuebeneaththe ridges. This pattern of hard tissue distribution is only found in the fin spines of some fossil chondrichthyans.Inacanthodians,thetissueinthefinspineridgesisorthodentineormesodentine, andisdifferentfromthecellularboneortrabeculardentinebeneaththeridges. 2. SinacanthfinspinesfromChinaandAustralia(TalentandSpencer-Jones 1963)alwayslackan inserted base. Thisis also thecaseformost sinacanths from Bolivia (Janvierand Suarez-Riglos 1986; Gagnier et al. 1988), and this feature is too common to be attributed to loss due to preservation.Theabsenceofabasemaybeplesiomorphicforchondrichthyans,bycomparisonwith theshortbaseinthefinspineofAntarctilamna(Young 1982).Thus,theshortbaseofinsertionin Sinacanthusboliviensis(Gagnieretal. 1988,fig. 8b-c)islikelytobederived,comparedwithother sinacanths. 3. Thefinspinesofsinacanthshavenumerousridges,alwaysmorethan 15persideandreaching upto 50perside(Zeng 1988). Alargenumberofridgesisalsocommonlyseenonthedorsalfin spinesofsharks(Maisey1981,1982;Young1982).Bycontrast,theridgesonacanthodianfinspines aregenerallyfewinnumber, althoughsomeclimatiidshaveupto 15 ridgesperside. Thisdistinctionpermitsreassessmentofsomeproblematicalfinspines.JanvierandSaurez-Riglos (1986)describedasmallfinspineassociatedwithascapulocoracoidfromtheuppermostSilurian ofBolivia. This fin spineis similarto Sinacanthus inits grossmorphology (Janvierand Saurez- Riglos 1986, fig. 5c), butcloseexaminationshowsthatithasfewerthan 15ridgesperside,asin otherclimatiidsandisthusdistinctfromsinacanths. Wangetal.(1980)erectedanewacanthodiangenus(Neoasiacanthus),representedbytwospecies fromtheFentouFormation(Wenlock)ofChaoxian,Anhui(Text-fig. 1). Pan(19866)considered thespinespecimensofN.wanzhongensisandN.shizikouensistobeplacodermspinalplates.Indeed, theholotypeofN.shizikouensis,inwhichtheposteriorfaceopensalmosttotheapexofthespine (Wangetal. 1980,pi.2,fig.9c),closelyresemblesaplacodermspinalplate,andisdistinctfromthe ZHU: EARLYSILURIANSINACANTHS 161 acanthodianfin spine. For this reason, Pan’s suggestion regarding the systematic position ofN. shizikouensisisfollowedhere. AstothetypespeciesofNeoasiacanthus(N. wanzhongensis), I still considerittobeanacanthodian. Thisspecies(Wangetal. 1980)hasfinspineswhicharesimilar tothoseofsinacanthsintheirgrossmorphologyandnodularridges(Text-fig.2g).However,since thesefinspineshaverelativelyfewridges(tentothirteenridgesperside),Neoasiacanthusismore similar to acanthodians than to sinacanths. Similar fin spines have also been found in western Hunan(Zeng 1988,finspinetype6), inassociationwithsinacanthfinspines. 4. Faunalassociationsalsosupportthechondrichthyanaffinityofsinacanths.InTarim,numerous chondrichthyanscaleswerefoundtogetherwiththesinacanthfinspines,butnoacanthodianscales were obtained. Until now, no microvertebrate remains have been collected from the sinacanth- bearingbedsofSouthChina.InBolivia,thesinacanthfinspineswerealsoassociatedcloselywith abundant chondrichthyan remains (Janvier and Suarez-Riglos 1986; Gagnier et al. 1988, 1989; Janvier 1991). The weight of evidence thus suggests that the sinacanths should be referred to the chondrichthyans, as provisionally suggested by Gagnier et al. (1988), rather than to the acanthodiansasoriginallyproposed. Itshouldbeemphasizedthatonlyfinspineswithmorethan 15ridgespersidecanbeassignedtosinacanths,whilethosewithfewerridgesarestillconsidered asacanthodians. Itmightbearguedthatsomesinacanthfinspines,suchasNeosinacanthus(P’an et al. 1975), are quite broad at the base and short, and resemble the intermediate fin spines of acanthodians. However, the ornamented part ofthe fin spine is also broad and short in some chondrichthyans(Maisey 1982). SYSTEMATIC PALAEONTOLOGY ClasschondrichthyesHuxley, 1880 Subclasselasmobranchii Bonaparte, 1838 Familysinacanthidaefam. nov. Diagnosis. Elasmobranchswithtapering,stronglycompresseddorsalfinspines;crosssectionfrom threeto sixtimesaslongaswide; anteriorfaceoffinspineacutelyrounded, lateralface slightly convexorflat,posteriorfaceconcaveandwithoutamedianridge;spinewallfairlythin,leavinga lnaurmgebecrentorfalricdagveistyin;csrpeianseessutrofwaacredmsatrhkeedbawsiethofmtohreestphinaena1s5alornegsiutlutdionfablirfiudrgceastiaonndagnrdoomvaersgi;ntahle insertion; ridgesgenerallyaswideasorwiderthangrooves; ridgeswithcloselysetorpectinated tubercles; tubercles of adjacent ridges never in contact; spine trunk and ridges composed of trabeculardentine;aninnerlayeroflamellardentineiswelldeveloped;noenamelonthesurface. Remarks. Sincethephylogenyofearlyelasmobranchsremainsobscure,thediagnosisgivenabove ismoredescriptivethanphylogenetic.Somecharactersarealsoseeninotherfinspinesandmaybe plesiomorphic.Themostcharacteristicfeaturesofsinacanthsincludethestronglycompressedfin spine, its thin wall and large central cavity, more than 15 longitudinal ridges per side and the marginalinsertionofridges. Sinacanthfinspinesresemblethoseoftheprimitiveelasmobranch Ctenacanthus(Maisey 1981) intheirnumerouslongitudinalridgesornamentedwithpectinatedtubercles.However,Ctenacanthus differsfromsinacanthsinhavingtuberclesofadjacentridgesalmosttouching(notinC.specablis), withamedianridgeontheposteriorface,andacrosssectiononlytwicetothreetimesaslongas itiswide.Themarginalinsertionofridgesisalsoabsent,butthisfeatureisseenalongtheposterior margininSphenacanthus(Maisey 1982).Thisformdiffersinthesmoothorwidelyspacednodular tuberculationoftheridges,andinhavingacrosssectiontwiceaslongasitiswide.Themarginal ridgeinsertionalongtheanteriormarginseeninsinacanthsisakindofbifurcation,aswasclearly showninAntarctilamna(Young 1982). 162 PALAEONTOLOGY,VOLUME41 GenussinacanthusP’an, 1959 Typespecies.SinacanthuswuchangensisP’an, 1959. Diagnosis.Sinacanthwithlongandslenderfinspines;spinegraduallytapering,recurvedposteriorly anddagger-shaped. SinacanthuswuchangensisP’an, 1959 Plate 1,figures2-3,6,8;Text-figure2a-e 1959 SinacanthuswuchangensisP’an,p. 11,pi.4,figs3-4. 1964 SinacanthuswuchangensisP’an,p.142,pi. 1,figs 1—4. pl973 SinacanthusfancunensisLiu,p.145,text-fig.2a,pi. 1,figs6-8(nontext-fig.2b,pi. 1,fig.5). 1988 Acanthodiiindet.finspine5,Zeng,p.291,text-fig.2f,pi. 1,figs 10-11. 1995 SinacanthuswuchangensisLiu,p.88,text-fig. 1a,pi. 1,figs 1-6. pl995 SinacanthusfancunensisLiu,p.89,text-fig. 1b,pi. 1,figs7-8(nontext-fig. lc,pi. 1,fig.9). Lectotype. Anincompletefinspine, MG.V1032a(Text-fig. 2a),designatedastheholotypebyP’an(1964); Wuhan,Hubei;GuodingshanFormation;Wenlock. Syntypes.Twooriginallyunnumberedfinspines(P’an1959,pi.4,figs3^1).One(P’an1959,pi.4,fig.3;Text- fig. 2a) later numbered as MG.V1032a, the other (P’an 1959, pi. 4, fig. 4; Text-fig. 2b) here numbered MG.V1032b. Referredspecimens.FinspinesdescribedbyP’an(1964,p. 143,pi. 1,figs2-4;Text-fig.2d-e)fromthetype locality;finspinesdescribedbyLiu(1973,p. 145,fig.2a,pi. 1,figs6-8;Text-fig.2c)fromNingguo(Anhui); finspinetype5,describedbyZeng(1988,pp.291-292,text-fig.2f,pi. 1,figs 10-11)fromwesternHunan; IVPP.V12093-12100, described by Liu (1995, pp. 88-89, text-fig. 1a-b, pi. 1, figs 1-8) from Tarim; IVPP.V11247.1-4(PI. 1,figs2-3,6,8)collectedbytheauthorandcolleaguesfromTarim. Diagnosis.SpeciesofSinacanthusinwhichthefinspineshaveashortbasalmarginandmorethan 15ridgesperside. Description. Asisclearlyshownbyitssyntypes(P’an 1959;Text-fig.2a-b),Sinacanthuswuchangensishasa fairlyslender,laterallycompressedandgentlytaperingfinspine,whoseridgesareornamentedbycloselyset tubercles,somewhatresemblingthoseofCtenacanthus.Theridgesincreasebasallybymeansofbifurcationand marginalinsertion.However,bothtypesarefragmentary.ThelectotypeMG.V1032a(Text-fig.2a)lacksthe EXPLANATIONOFPLATE 1 Figs1,4.Neosinacanthussp.2;IVPP.V11249;TataaiertageFormation(Llandovery);Kalpin,Xinjiang,China. 1,crosssectionofthefinspine; x60.4,afinspine; x2.Abbreviations:lam,lamellardentine;tra,trabecular dentine. Figs2-3,6,8.SinacanthuswuchangensisP’an, 1959.2,IVPP.VI1247-1;anelastomerecastofthefinspine, showing bifurcation and marginal insertion of ridges; Tataaiertage Formation (Llandovery); Kalpin, XBiancjhiua,ngX;injxi5a.ng3,;IxVP2.P.6,V1IV1P2P47.-V3I;1a2n47e-l4a;staonmeirnetecransatlocfasttheoffitnhespciennet;raYlicmauvgitayntofawaufiFnosrpmianet;ioYnim(uWegnalnotcakw)u; Formation(Wenlock);Bachu,Xinjiang; x2.8,IVPP.VI1247-1;SEMphotographofanelastomerecastof the fin spine, showing ornamentation; Tataaiertage Formation (Llandovery); Kalpin, Xinjiang, China; x30. FigF.o5r.mNaetoisoinna(cLalnatnhduosveprlyan)i;spKianlaptiuns,PX’iannjiaanndg,LiCuh,iinnaP;’axn2.etal.,1975;IVPP.V11248;afinspine;Tataaiertage Fig. 7. Sinacanth; IVPP.VI1252; SEM photograph of fin spine fragment, showing ornamentation; YimugantawuFormation(Wenlock);Bachu,Xinjiang,China; x45. Figs9-10. Tarimacanthusbachuensisgen.etsp. nov.;IVPP.VI1250;afinspine;YimugantawuFormation (Wenlock);Bachu,Xinjiang,China; x2. PLATE 1 ZHU, sinacanths ; text-fig. 2. a-e, reconstruction of the fin spine of Sinacanthus wuchangensis P’an, 1959. A, lectotype, MG.V1032a (P’an 1959, pi. 4, fig. 3). b, one ofthe syntypes, MG.V1032b (P’an 1959, pi. 4, fig. 4). c, IVPP.V4412a(Liu 1973,pi. 1,fig.7).d,MG.V1033(P’an 1964,pi. 1,fig.2a).E,MG.V1036(P’an 1964,pi. M1,Gf.igV.13)5.0F1,re(cPo’nasntreutcatli.on19o7f5t,hepih.ol1o0t,yfpieg.fi4n).spGi,neroefcoSnisntarcuacnttihounsotfritahnegulhaotluostPy’paenfainndspLiinue,ionfPN’eaonaestiaalc.,an1t9h7u5s wanzhongensisWang,XiaandChen, 1980;VF0262(Wangetal. 1980,pi.2,fig.6).H,reconstructionofthe finspineofSinacanthussp.;HV006.1 (Zeng 1988,pi. 1,fig.6).Scalebarsrepresent5mm. ZHU: EARLYSILURIANSINACANTHS 165 apical and basal extremities, and MG.V1032b consists only ofthe apical part (Text-fig. 2b). P’an (1964) assigned some morecomplete fin spines from the type locality to S. wuchangensis, and supplemented the original diagnosis. MG.VI033 (Text-fig. 2d), which was incorrectly designated as the paratype of S. wuchangensis,showstheadditionofridgesbybifurcationandmarginalinsertion.MG.VI036(Text-fig.2e)is acompletefinspine,showingtherelativelyshortbasalmarginofS.wuchangensis.IVPP.V4412a(Text-fig.2c) isoneofthesyntypesofS.fancunensis(Liu 1973).Asmentionedabove,therearenosignificantdifferences betweenthetypesofS.wuchangensisandtheslenderfinspinesofS.fancunensis. ManyfinspinesofS.wuchangensisfromTarimarewellpreservedandfairlycomplete(PI. 1,figs2-3).They alllacktheinsertionbase.Theanglebetweentheanteriorandbasalmarginsrangesfrom50°to90°.Thefin spinesareslender,laterallycompressed,andrecurvedposteriorlyattheapex.Thefinspinewallisthinand encloses a relatively large central cavity (PI. 1, fig. 6). The ornamentation of ridges is well shown in IVPP.V11247(PI. 1,figs2,8),andissameasthatofthesyntypes.SEMmicrographsofsomefragmentsof sinacanthsfromTarimshowdetailofthetuberclesonthesurfaceofridges.Theornamentation(PI. 1,fig.7) issimilartothatseeninsomespeciesofCtenacanthus,suchasC.littoni(Maisey 1981). Remarks. Liu (1973) established Sinacanthusfancunensis for fin spines from Anhui, but did not assignaholotype,soallthespecimensinhisstudymustberegardedassyntypesofS.fancunensis. Liu(1973)acknowledgedthatthisspecieswasverysimilartothetypespeciesofSinacanthus,but proposedtwodifferences: theanteriorandposteriormarginsofthefinspineofS.fancunensisare nearlyparallelinthelowerhalf,whereasthoseofS. wuchangensisconvergegentlyoverthewhole spine,andtherearec.25-30ridgespersideinthemiddlepartofthefinspineofS.fancunensis,but lessthan20inS.wuchangensis.NeitherdistinctionwaslatersupportedbyLiu(1995),whoproposed twoalternativedifferencesbetweenthesespecies:(1)thefinspineofS.fancunensisismoreslender thanthatofS. wuchangensis-,(2)theanglebetweentheanteriorandbasalmarginsisabout70°in S. wuchangensisandcloseto90°inS.fancunensis. However, thesetwodifferencesaredifficultto applyinpractice, and are inmy opinion invalid. Somefin spines referred to S. wuchangensisby Liu (1995), such as IVPP.V12094 (Liu 1995, pi. 1, fig. 2), are as slender as those referred to S. fancunensisbyLiu(1973, 1995). Inaddition,theanglebetweentheanteriorandbasalmarginsof thefinspinerangesbetween70°and90°,andismoreprobablyduetoindividualvariationrather than any specific significance. There are no definite differences between the slenderfin spines of S.fancunensisandS.wuchangensis,henceS.fancunensisshouldberegardedasajuniorsynonymof S.wuchangensis.The‘intermediatefinspines’ofS.fancunensis(Liu1973, 1995)shouldbereferred toanewsinacanth(asdiscussedbelow). FinspinesofS.boliviensis(Gagnieretal. 1988)arequitesimilartothoseofS. wuchangensisin outlineandornamentation,butdifferinthepossessionofashortinsertionbase. The Australian specimens referred to as ‘cf. Sinacanthus' by Turner (1986) much resemble S. wuchangensisintheirslendershapeandnumerousnodularridges(TalentandSpencer-Jones 1963). Theyalsolackaninsertionbase,assuggestedbyTurner(1986).TheydifferfromS. wuchangensis intheirstraighterconfigurationandintheadditionofridgesbyintercalationaswellasbifurcation andmarginalridgeinsertion.Therefore,theyprobablybelongtoadifferentspeciesofSinacanthus, S.micracanthus(Chapman, 1917). SinacanthustriangulatusP’anand Liu, inP’anetal., 1975 Text-figure2f 1975 SinacanthustriangulatusP’anandLiu,inP’anetal.,p. 164,pi. 10,fig.4. Holotype. MG.V1501, a complete fin spine (P’an etal. 1975, pi. 10, fig. 4; Text-fig. 2f); Wuhan, Hubei; GuodingshanFormation;Wenlock. Diagnosis.SpeciesofSinacanthusinwhichthefinspinehasaverybroadbasalmarginformingan acuteangle(about30°)withtheanteriormargin.Thereareupto50ridgespersidenearthebase. Remarks. ThisspeciesisreferredtoSinacanthusbecauseoftheslenderandrecurvedshapeofthe finspine(Text-fig. 2f). ItdiffersfromotherspeciesofSinacanthusintheacuteanglebetweenthe 166 PALAEONTOLOGY,VOLUME41 anteriorandbasalmarginsofthefinspine.Inaddition,thebasalmarginisaslongastheposterior marginofthespine. Sinacanthussp. Text-figure2h 1988 Acanthodiiindet.finspine4,Zeng,p.290,text-fig.2d,pi. 1,fig.6. Referredspecimens. Fin spines describedby Zeng(1988) as ‘fin spines4’, HV006.1-26; Dayong, Hunan; RongxiFormation;Llandovery. Remarks. Thiskindoffinspine(Text-fig.2h)representsanewspeciesofSinacanthus,diagnosedas ‘Sinacanthuswithafinspinewhosebasalextremityisanteroposteriorlyextended’.Thesefinspines exhibitthemaincharactersofSinacanthus'.numerousnodularridges(30-35ridgespersideatthe baseofthefinspine), theadditionofridgesbymarginalinsertionandbifurcation,andaslender andposteriorlyrecurvedconfiguration.TheydifferfromS. triangulatusinthatthebasalmarginis shorterthantheposteriormargin.Inaddition,theanglebetweentheanteriorandbasalmarginsin thisunnamedspeciesisabout45°,andthuslargerthanthatinS.triangulatuswhereitisabout30°. GenusneosinacanthusP’anandLiu, inP’anetal., 1975 Typespecies.NeosinacanthusplanispinatusP’anandLiu,inP’anetal. 1975. Diagnosis. Sinacanthwithbroadfinspine; anteriorandposteriormarginsarestraight. Remarks. Thisgenusisdistinguishedfromothersinacanthsbythestraightanteriorandposterior marginsofthefinspine. NeosinacanthusplanispinatusP’anand Liu, inP’anetal., 1975 Plate 1,figure5;Text-figure3a,c 1975 NeosinacanthusplanispinatusP’anandLiu,inP’anetal.,p. 165,pi. 10,fig. 5a-b. 1988 Acanthodiiindet.finspine2,Zeng,p.290,text-fig.2b,pi. 1,fig. 13. Holotype. MG.V1502,acompletefinspine(P’anetal., 1975,pi. 10,fig.5a-b;Text-fig.3a);Wuhan,Hubei; GuodingshanFormation;Wenlock. Referredspecimens. FinspinesdescribedbyZengas‘finspines2’fromtheRongxiFormation(Llandovery) ofDayong,HV005.1-3(Text-fig.3c);afinspinefromTarim,IVPP.V11248(PI. 1,fig.5). Diagnosis. SpeciesofNeosinacanthusinwhichthefinspineiswiderthanlong; theanglebetween anteriorandposteriormarginsisgreaterthan70°; andtheposteriormarginisarmedwithsmall triangulardenticles. Remarks. ThespecimensfromHunan(Text-fig.3c)andXinjiang(PI. 1,fig.5)areslightlydifferent fromtheholotype(Text-fig.3a).Theanglebetweentheanteriorandposteriormarginsreaches90° intheholotype,andisevidentlylargerthanthatintheHunanandXinjiangspecimens. However, this difference may reflect individual variation. The discovery ofNeosinacanthusplanispinatusin Xinjiang(Tarim)representsanotherelementcommontotheTarimandSouthChinafaunasduring theSilurian(cf. Liu 1993). Neosinacanthussp. 1 Text-figure3b 1988 Acanthodiiindet.finspine 1,Zeng,p.289,text-fig.2a,pi. 1,fig. 1.